Models and tests of anchorage zones of posttensioning tendons

Autor(en):

Burdet, Olivier L. / Sanders, David H. / Roberts, Carin L.

Objekttyp:

Article

Zeitschrift:

IABSE reports = Rapports AIPC = IVBH Berichte

Band (Jahr): 62 (1991)

PDF erstellt am:

03.11.2015

Persistenter Link: http://dx.doi.org/10.5169/seals-47684

Nutzungsbedingungen
Mit dem Zugriff auf den vorliegenden Inhalt gelten die Nutzungsbedingungen als akzeptiert.
Die ETH-Bibliothek ist Anbieterin der digitalisierten Zeitschriften. Sie besitzt keine Urheberrechte an
den Inhalten der Zeitschriften. Die Rechte liegen in der Regel bei den Herausgebern.
Die angebotenen Dokumente stehen fr nicht-kommerzielle Zwecke in Lehre und Forschung sowie fr
die private Nutzung frei zur Verfgung. Einzelne Dateien oder Ausdrucke aus diesem Angebot knnen
zusammen mit diesen Nutzungshinweisen und unter deren Einhaltung weitergegeben werden.
Das Verffentlichen von Bildern in Print- und Online-Publikationen ist nur mit vorheriger Genehmigung
der Rechteinhaber erlaubt. Die Speicherung von Teilen des elektronischen Angebots auf anderen
Servern bedarf ebenfalls des schriftlichen Einverstndnisses der Rechteinhaber.
Haftungsausschluss
Alle Angaben erfolgen ohne Gewhr fr Vollstndigkeit oder Richtigkeit. Es wird keine Haftung
bernommen fr Schden durch die Verwendung von Informationen aus diesem Online-Angebot oder
durch das Fehlen von Informationen. Dies gilt auch fr Inhalte Dritter, die ber dieses Angebot
zugnglich sind.

Ein Dienst der ETH-Bibliothek

ETH Zrich, Rmistrasse 101, 8092 Zrich, Schweiz, www.library.ethz.ch
http://retro.seals.ch

545

Models and Tests of Anchorage Zones of Post-Tensioning Tendons

Modeles et essais de zones d'ancrage des cables de precontrainte

Modelle und Versuche von Verankerungszonen von Vorspannkabeln

Olivier

L. BRDET

Scientific Associate
Fed Inst of Technology
Lausanne, Switzerland

Carin L. ROBERTS
Ph D candidate
Univ of Texas
Austin, TX, USA

David H. SANDERS
Prof of Civil Eng
Univ of Nevada
Reno, NV, USA

John

E. BREEN
Prof of Eng
Univ of Texas at Austin
Austin, TX, USA

Gregory L. FENVES
Prof of Civil Eng
Univ of California
Berkeley, CA, USA

SUMMARY
This paper presents the results of an investigation of the behaviour and the design of anchorage
zones of post-tensioning tendons The analytical component is a combination of Finite Element
Analysis and Strut-and-Tie Models A total of more than 60 tests of anchorage zones are included
in discussion and practical guidelines for the design proposed for incorporation in the AASHTO
Bridge Design Specification are outlined
RESUME
Cet article presente les resultats d'un projet de recherche sur le comportement et le
dimensionnement des zones d'ancrage des cables de precontrainte La partie analytique comprend la fois une analyse par la methode des elements finis et des modeles de treillis Au total,
cet article inclut les resultats de plus de 60 tests expenmentaux de zones d'ancrage et inclut des
directives pratiques qui ont ete proposees pour etre inclues dans la norme amencaine de ponts
routiers AASHTO

ZUSAMMENFASSUNG
Im vorliegenden Bericht werden die Resultate eines Forschungsprojektes ber das Verhalten und
die Bemessung von Verankerungszonen von Vorspannkabeln beschrieben Der analytische Teil
beinhaltet sowohl Finite Element Berechnungen als auch Fachwerkmodelle Die Resultate von
mehr als 60 Versuchen an Verankerungszonen werden aufgefhrt Weiter enthalt dieser Bericht
praktische Richtlinien, die fr die Aufnahme in die amerikanische Strassenbrucken-Norm
AASHTO vorgeschlagen wurden

546

1.

MODELS AND TESTS OF ANCHORAGE

Introduction

The quest for development of a consistent approach to

concrete clearly requires a hierarchy of highly transpare
oriented analysis tools [2]. These will rnge from
traditional section mechanics principles suitable fo
B-regions to the more intuitive strut-and-tie models
more formal elastic or non-linear finite element analys
required for the D-regions. Scordelis [12] indicates that
latter are extremely useful, "... it is imperative that ex
and qualified structural engineers be involved
interpretation of the results using their judgem
knowledge of structural behavior..." MacGregor [5] reite
need but gives special emphasis in the D-regions sayi
details of the reinforcement in the discontinuities co
strength of these regions and hence must be consider
structural engineer." Marti [6] suggests that in usual ap
of STM, the design is rather insensitive to the assessm
effective concrete stress, fc. While this is true
applications, it is clearly not true in design of postanchorage zones. In such discontinuity zones, the v
forces transmitted to the concrete by the tendon a
cause very high local stresses on the concrete. The sp
these forces through the member causes substantia
serviceability limit State, with undesirable cracking, and
the anchorage zone need to be prevented.
Test results and failures during construction (See Fig
nodes or at the intersection of confined nodes and un
zones. This particular detailing application thus poses
methods since assessing node and strut capacity is far
dimensioning of reinforcement.

O.L BRDET, D.H.

SANDERS,

CL

ROBERTS, J E BREEN, G

547

FENVES

Local compression failure, in which the failure occurs at a very Short distance from the anchorage device, and is
caused by lack of confinement in the area immediately surrounding the anchorage device.
Compression failure, similar to the previous mode failure, but with the difference that the failure occurs at a
larger distance from the anchorage device, which is itself sufficiently confined.
Tension failure, in which the reinforcement provided to resist the tensile force induced by the spreading of the
concentrated tendon load is insufficient.
-General Zone

two regions in the anchorage zone. The local

zone, in the immediate vicinity of the anchorage device, is highly
dependant on the post-tensioning system and is the
responsibility of the supplier of the anchorage device. The
general zone is more remote from the anchorage device and is
less influenced by the post-tensioning system. It is the
responsibility of the structural engineer. Of the three modes of
failure described above, the first one occurs in the local zone,
the second mode of failure occurs in the general zone, most
often at the interface with the local zone, and the third mode of
failure occurs in the general zone.
In order for anchorage devices to be deemed satisfactory, they
need to either meet maximum bearing stress and minimum
stiffness requirements or to be tested following a prescribed
testing procedure described in Section 4. The distinction between
local and general zone gives flexibility to the constructor, who
can choose the anchorage device and the post-tensioning system,
without jeopardizing the integrity of the structure, and without Figure 3:
unduly complicating the work of the design engineer.
Figure 3 shows

4. Local Zone Tests

to 1.5h
a) Elevation

General Zone

Ha

b) Plan view

^^ j^^

a-a

and General Zone

by Roberts

A part of the NCHRP Anchorage Zone research project consisted in an investigation focusing on the behavior of
local anchorage zones both at service State and at ultimate. The purpose of this study by Roberts [9] was to define
the test procedures and compliance criteria for the testing of anchorage devices. Roberts tested 31 local zone test
specimens. The behavior of local anchorage zones was found to be sensitive to the type and amount of confining
reinforcement, as well as to the cover provided around the anchorage device. Existing formulae by Richart [8] and
Nyogi [7] were enhanced to give a better prediction of the strength of a local zone. Cyclic testing of local zones
gives results similar to extended (48 hours) testing, and is more representative of the behavior of anchorage zones
under field conditions than monotonic testing. A standardized testing procedure for the local zone was proposed
by Roberts for introduction in the AASHTO Bridge Specification.
5

Finite Element Analysis and Strut-and-Tie Models

7Fnu
zone configurations, therefore the design of the
A'/*
general zone must be approached in a different manner than the local zone. The number of
*A + + + -Y
variables affecting the design of the anchorage zone remains large even though the local zone
has been addressed. A survey of the current design practice in the United States by Sanders
[10] showed that the post-tensioning industry is very creative. Tendons often present an
eccentricity, an inclination and a curvature
in the anchorage zone. Multiple tendons
are commonly used, in groups of two to six
tendons. Transverse post-tensioning and
0 25.
+ Finite Element
transverse reactions are often present in the
Guyon Analysis
OZ
zone. Special geometries are Figure 4: Simple
anchorage
Bursting
Approx mote
Force
used to introduce the post-tensioning force Strut-and-Tie
Tours
Formulos
to the section, using for example blisters or Model with
o 0 3P(1 -o/h)
ribs. The first phase did not consider the Elastic Stress
-o 0 25P( 1-o/h)
0 05.
expanding field of external post-tensioning. Vectors
The project was set up to use a
combination of elastic finite element analysis, strut-and-tie
0 0102 0 3 04 05 060 7 0809
models and physical tests. Linear elastic finite element analysis
piote
s.ze
o/h
Reiot^
offers the advantage of being a well known method of obtaining
Figure 5: Bursting Force for Concentric
the internal state of stresses in a body. As pointed out by
Tendons

It is nor practical to test all possible general

548

MODELS AND TESTS OF ANCHORAGE ZONES OF POST-TENSIONING TENDONS

Ties

1
aT
a
(75MU-IU5_a20 25 03 035

>*-a
0

Force in

Struts ond

73

Froction
of P

Struts

-0

t>)

Relative Eccentricity

'P3

Strut

e/h

Figure 6: Strut-and-Tie Figure 7: Forces in the Struts and Ties

Model of an Eccentric Function of the Eccentricity
Anchorage Zone

Figure 8: Strut-and-Tie Model of an

Anchorage Zone with a Curved
Tendon, Eccentricity 0.25h and Initial
Inclination 20 degrees
Schlaich [11], the elastic State of stresses constitutes a good starting point for the development of strut-and-tie
models. Of special interest is the representation of the principal stress vectors shown in Figure 4. These vectors
give a good idea of the flow of forces through the anchorage zone and are helpful in assessing the adequacy of a
strut-and-tie model. The physical test specimens by Sanders [10] were used to demonstrate the validity of the
models and to calibrate the design formulae.
Figure 5 shows the bursting force obtained by integrating the elastic stresses perpendicular to the tendon path,
along with the force obtained from the simple strut-and-tie model shown in Figure 4. As can be observed, the
correlation is quite good. The figure also shows Guyon's equation [4] for the same force.
The real power of the strut-and-tie model is its ability to model a
wide rnge of anchorage zone configurations. Figure 6 shows a strutand-tie model for an anchorage zone with an eccentric load. Figure 7
rtOutxJe
shows the forces in the various members as a function of the
0 rt
Tronsverse
eccentricity.
1
Bort)
F.0
Force
f-F() <V
Figure 8 shows two possible strut-and-tie models for an anchorage
A hopwtd fwmdo
zone with an eccentricity, an inclination and a curvature of the
o/hWo.*
tendon. If no tie-back reinforcement is provided, all the tendon
deviation forces are transmitted to the strut on the inside of the
tendon, and the external strut is straight between the reinforcement
bars. If tie-back reinforcement is provided, the tendon deviation
Initiol Incbnotton ir
forces are distributed to both compression struts. If the force in the
tie-back reinforcement is added to the bursting force, it is found that
Figure 9: Transverse Bursting Force as a
the sum ^ approximately equal to the bursting force in the case
Function of the Initial Inclination for an
without tie-back reinforcement. Figure 9 shows the Variation of the
Initial Eccentricity of 0.25h
bursting force as a function of the initial inclination. The figure also
shows the results obtained from the finite element analysis and the values predicted by an approximate formula as
outlined in Section 6. In a simplified fashion, the increase in tensile force caused by the inclination of the tendon
can be approximated as one half of the net shear on the general zone summing the effect of external loads and the
transverse component of the post-tensioning force. This corresponds to the intuitive idea that roughly half of the
force is resisted by each compression strut.

^y^^

i^

Q 3

as a

lue)

(0

STM

O- STU

Back)

0.2WO

0 175.

SB

\- 1-0 1!

-0 12P

CWD

Q- Tension on Tendon Axis

0 150.

from

0125.

--

quorter points) ond EF/GH

(outside quorter points)
from Strut-and-Tie Model

0100.
0075.

Ar Tension on Top of
Centerline from FE

0 050.

TT

0I5P
within the
Quarter Points
(s
0175h)

a) tx>ad

Figure 10:
Examples of
Models with two Tendons

35P

TT

0 35P

0 15P

b) U>ad Outside the

Quarter Points
0 35h)
(s

Strut-and-Tie

FE

Tension in Tie GH (within

0 025.

- Tension
0 000.

ttf&^

Holt Oistonce Between Anchors

^"

in Top Tie BC

(outside quorter points)

from Strut-and-Tie Model

s/h

Figure 11: Tension Forces in an Anchorage Zone

with Two Tendons

O. L. BRDET, D. H. SANDERS, C. L ROBERTS, J. E. BREEN, G. L FENVES

549

The effect of groups of tendons was investigated, and it was found that, for rectangular sections with straight
tendons, the largest tensile forces are induced when only two tendons are used. Figure 10 shows the two basic
configurations for two anchorages. If both tendons act within the kern of the section, the State of stresses is
similar to that induced by a single anchorage device. As the tendons move outside the kern, an increasingly large
edge tension force is induced between the anchorage devices close to the surface of the concrete. Figure 11 shows
the edge tension force between the anchorages as a function of the spacing of the anchorages.
Because the compressive stresses in the immediate vicinity of the anchorage device are usually higher than the
uniaxial compressive strength of the concrete fc, the verification of the capacity of the concrete compression struts
is critical. Most authors assume that the limiting stress in the concrete struts is some fraction of fc, which is too
constraining for anchorage zones. Sanders [10] incorporated the effect of confining reinforcement in a strut-andtie model. For most practical cases, however, the checks involved in such calculations are beyond the capabilities
of the engineer. Therefore, the local zone acceptance tests are relied on for determining the adequacy of the
confined node. The critical section for the verification of the compressive stresses in the concrete struts is defined
at a certain distance from the anchorage device (in general at one times the lateral dimension of the anchorage

device). This allows one to check the compressive stresses in the concrete against the conventional value
which is commonly aeeepted for strut-and-tie models.
6. Design Method

of 0.10fc

for Anchoraee Zones

The goal of the NCHRP Anchorage Zones research project is the elaboration of a clear, consistent and easily
applicable method for the design of anchorage zones of post-tensioning cables. Assuming that the engineer has a
good knowledge of the location and magnitude of the force for each tendon, some idea of the size of the
anchorage device that will be required to transmit the force and the assurance that the anchorage device used
satisfies the testing requirements of Section 4, guidelines for the design of the general anchorage zone are needed.
A number of procedures are suggested in the proposed AASHTO revisions. Two general procedures are allowed.
One is a detailed elastic analysis such as a valid finite element analysis (FEA). Rules are provided for integrating
tensile stresses and selecting appropriate limiting stress values. The second procedure allowed is the strut-and-tie
model (STM). Since this equilibrium based procedure is not sensitive to compatibility induced stresses at service
load levels, such as edge tension, or spalling stresses around anchorages, certain guidelines are provided requiring
supplemental spalling crack control reinforcement. Recognizing that either FEA or STM solutions may require
considerable extra effort for the design of some relatively simple but common applications, an approximate
procedure is also included. This procedure was developed from the results of FEA and STM parametric
studies [3]. It uses relatively simple formulae to determine the magnitude and location of the bursting force and
to check the compressive stress at the interface between the local zone and the general zone. It is limited to the
case of a single anchorage, or of a Single group of closely spaced anchorages acting on a rectangular cross section.
7. Evaluation

of the Methodology based on Test Results by Sanders

Sanders [10] condueted a series of 36 tests of anchorage zones. In the specimens modelling Single tendon
anchorage zones, the reinforcement patterns and the tendon eccentricity, inclination and curvature were varied.
Tests of anchorage zones with multiple tendons were also condueted, with the prime variable being the spacing
between the anchors. The cracking load of 31 of the specimens by Sanders was estimated based on the elastic
stress distribution obtained from a two-dimensional Finite Element Analysis. and the tensile strength of the
concrete measured from split-cylinder tests. The average ratio of actual to predicted cracking load is 0.91, with a
Standard deviation of 0.22. Figure 12 shows the ultimate load reached by the same series of specimens, along with
the ultimate load predicted using strut-and-tie models based on elastic stress resultants at the end of the general
zone. The average ratio of predicted to ultimate is 1.44, with a Standard deviation of 0.44. Sanders [10] developed
enhancements to the cracking load prediction, including the effect of the reduction of the tensile strength of the
concrete caused by the three-dimensional state of stresses in the anchorage zone. Taking this modification into
account, the average ratio of actual to predicted cracking load becomes 1.05 for all tests, with a Standard deviation
of 0.20. For the ultimate load, Sanders also developed an enhanced STM which includes the effect of a limited
plastification of the concrete in the immediate vicinity of the anchorage device. Taking this modification of the
model into account brings the average ratio of the predicted ultimate load to the actual ultimate load to 1.19, with
a Standard deviation of 0.19.
One of the most notable observations made during the evaluation of the test results is the fact that in the large
majority of the cases, the capacity of the anchorage zone is controlled by the strength of the compression struts at
the interface between the local zone and the general zone. At this location the concrete has no confinement, and
is exposed to very large compressive stresses. Thus, increasing the reinforcement of the general anchorage zone
will in many case lead to little or no improvement of the overall strength of the anchorage zone. This is confirmed
by the Observation of Stone and Breen [13], who noted that increasing the amount of orthogonal reinforcement

550

MODELS AND TESTS OF ANCHORAGE ZONES OF POST-TENSIONING TENDONS

(the reinforcement provided in the

general zone) is not nearly as effective as
1000
using longer and heavier spirals, which
confine the local zone and have the effect
of displacing the interface between the
local zone and the general zone to an Ultimate
Load
area of lower compressive stresses. For in kips
design purposes, it is in any case
advisable to remember that the stresses
in the concrete struts often control the
design. Also notable is the effect of
tensile stresses existing in the anchorage
TP TP TP TP
M1M2M2M3M3M4M4M5M6ME
B1B2B3B4B5B6B7B8 E1E2E3E4E5E6M5
zone. The resistance these stresses
T1T2T3T4
is
the
design,
provide usually neglected in
Actual Ultimate Load
but it nevertheless plays an important
g STM Predicted Ultimate Load
role in the behavior of anchorage zones.
In several cases it was observed that the Figure 12: Actual Ultimate Load and Predicted Ultimate Load for
strength of the anchorage zone exceeded Sanders' Test Specimens (1 kip 4.54 kN)
that predicted based on the capacity of
the tension ties alone. Brdet [3] suggests that this additional strength is caused by the fact that a part of the
concrete at the base of the specimens remained uncracked up to failure, thus providing an additional tensile
capacity to resist bursting forces.
11

12 13 14

8, Conclusions

The analysis, behavior and design of anchorage zones of post-tensioning tendons was investigated using a
combination of Finite Element Analysis, Strut-and-Tie Models and experimental test specimens. This
combination allowed minimization of the number of required experimental specimens and generalization the
results in the form of simple design formulae. A consistent design methodology allowing use of finite element
analyses, strut-and-tie models, and for certain frequently occurring cases, relatively simple design formulae was
developed and has been proposed for inclusion in the AASHTO Bridge Design Specification. A Standard testing
procedure for anchorage devices and their necessary confinement was also proposed.
The cracking loads computed based on the elastic stresses and the split cylinder strength of the concrete are
slightly smaller than the actual cracking loads, possibly because of the detrimental effect of the transverse
compression. The ultimate capacity of anchorage zones can be conservatively predicted using the Strut-and-Tie
Model. This investigation clearly indicates the critical nature of the compressive struts in anchorage zones. This
differs from many other D-region applications in which the struts are not as criotical.
References

[I]
[2]
[3]
[4]
[5]

[6]
[7]

American Association of State Highway Transportation Officials (AASHTO),"Standard Specification for Highway Bridges",
13th edition, 1983.
Breen, J.E., "Why Structural Concrete," Final Report IABSE Colloquium on Structural Concrete, Stuttgart, April 1991.
Brdet, O.L., "Analysis and Design of Anchorage Zones in Post-Tensioned Concrete Bridges," PhD dissertation, University
of Texas at Austin, May 1990.
Guyon, Y., "Prestressed Concrete," John Wiley and Sons, New York, 1953
MacGregor, J.G., "Sub-Theme 2.4 - Dimensioning and Detailing," Final Report IABSE Colloquium on Structural Concrete,
Stuttgart, April 1991.
Marti, P., "Sub-Theme 2.4 - Dimensioning and Detailing," Final Report IABSE Colloquium on Structural Concrete,
Stuttgart, April 1991.
Nyiogi, S.K., "Bearing Strength of Reinforced Concrete Blocks," ASCE Structural Division Journal, Vol. 101, No ST5, May
1975.

[8] Richart, F.E., Brandtzaeg, A, and Brown, R.L., "A Study of the Failure of Concrete under Combined Compressive
Stresses," Research Bulletin No 185, University of Illinois Engineering Experimental Station, 1928
[9] Roberts, C.L., "Behavior and Design of the Local Anchorage Zone in Post-Tensioned Concrete," MS thesis, University of
Texas at Austin, May 1990.
[10] Sanders, D.H., "Design and Behavior of Post-Tensioned Concrete Anchorage Zones," PhD dissertation, University of Texas
at Austin, August 1990.
[II] Schlaich, J., Schfer, K., Jennewein, M., "Towards a Consistent Design of Structural Concrete," PCI Journal, Vol. 32, No 3,
May-June 1987, pp 74-151.
[12] Scordelis, AC, "Analysis of Structural Concrete Systems," Final Report IABSE Colloquium on Structural Concrete,
Stuttgart, Apnl 1991.
[13] Stone, W.C., Breen, J.E., "Behavior of Post-Tensioned Girder Anchorage Zones," Center for Transportation Research
Report No 208-2, University of Texas at Austin, January 1981.